Wednesday, August 30, 2006

After Revelations of Undisclosed Conflicts of Interest, a Journal Editor Steps Down

In July, we briefly posted about several cases in which articles published in prominent journals were discovered to have been written by authors who had relevant, but undisclosed conflicts of interest. One of those cases we summarized thus:
  • An article was published in Neuropyschopharmacology in July about vagus nerve stimulation as a treatment for depression [Nemeroff CB, Mayberg HS, Krahl SE. VNS therapy in treatment-resistant depression: clinical evidence and neurobiological mechanisms. Neuropsychopharmacology 2006; 31, 1345–1355.] The Wall Street Journal discovered that eight of the article's nine authors had financial ties to Cyberonics Inc, the manufacturer of the device. The ninth author is an employee of the company, which was disclosed." (See previous post on Cyberonics here.)
In August, Bernard Carroll posted in more detail. He noted that the case involved more than the failure to disclose possible conflicts of interest, but also that one of the authors of the first draft the article was a hired writer, whose role was not fully disclosed; that "the review carefully followed the corporation's marketing message and branding language"; and that Cyberonics produced a press release touting VNS therapy featuring Nemeroff, but without revealing his financial relationships with the company.

Now the Wall Street Journal, and The Scientist reported that Nemeroff, who not only wrote the article in question, but was the Editor of Neuropsychopharmacology, will be stepping down from that role in December, "in part, based on the recent adverse publicity to the journal and the ANCP [American College of Neuropsychopharmacology]."

The Scientist also noted that "This isn't the first time that Nemeroff has hit the headlines for undisclosed financial ties. In 2003, a review he coauthored in Nature Neuroscience neglected to mention significant financial interests in three therapies that were reviewed favorably (including owning the patent on one of the treatments), prompting the Nature Publishing Group to widen its disclosure policies."

The Scientist quoted Clare Stanford, past President of the British Association for Psychopharmacology, "I don't believe for a minute that the fact the paper was funded by a company would have influenced his conclusions. It is unfortunate that he has had to stand down over this incident which is largely a reflection of the scientific community's paranoia rather than any failing of his professional integrity."

Is this all paranoia? We should note that cognitive psychological research suggests that even small gifts affect peoples' cognition. Research to this effect was one motivation for a widely publicized call for physicians to reject even gifts like pens and coffee mugs (see post here). If such small gifts to practicing physicians are so worrisome, what about more substantial conflicts affecting clinical researchers, or medical journals? Of course, the effects of conflicts of interest on judgments and decisions are often unconscious, so that those with conflicts really do believe that these conflicts do not affect them, perhaps explaining some of the resistance to even disclose conflicts. (See a summary from the Wall Street Journal via the Pittsburgh Post-Gazette.)

However, research summarized by Carl Elliott in the Journal of General Internal Medicine (see post here) suggested that disclosure is not the answer, and may make things worse. More cognitive psychological research suggested that people discount the amount of bias caused by conflicts of interest, and that disclosure may cause those with conflicts to be even more biased. Thus Elliot wrote, "the solution to the bias created by conflict of interest is not simply to disclose the conflict, which makes bias even worse. Rather, the solution is to eliminate the financial conflicts."

Amen.

No comments: